[Bug][MoE] Fix TRTLLM NVFP4 Routing Kernel Precision (#36725)

Signed-off-by: Robert Shaw <robshaw@redhat.com>
Co-authored-by: Robert Shaw <robshaw@redhat.com>

vllm/model_executor/layers/fused_moe/experts/trtllm_nvfp4_moe.py

@@ -298,10 +298,7 @@ class TrtLlmNvFp4ExpertsMonolithic(
             and self.routing_method_type != RoutingMethodType.Llama4
         )
 
-        # Prepare routing bias into kernel format.
-        routing_bias = e_score_correction_bias
-        if routing_bias is not None:
-            routing_bias = routing_bias.to(torch.bfloat16)
+        # Prepare router logits for kernel format.
         router_logits = (
             router_logits.to(torch.float32)
             if self.routing_method_type == RoutingMethodType.DeepSeekV3
@@ -311,7 +308,7 @@ class TrtLlmNvFp4ExpertsMonolithic(
         # Invoke kernel.
         return flashinfer.fused_moe.trtllm_fp4_block_scale_moe(
             routing_logits=router_logits,
-            routing_bias=routing_bias,
+            routing_bias=e_score_correction_bias,
             hidden_states=hidden_states,
             hidden_states_scale=a1q_scale.view(torch.float8_e4m3fn).reshape(
                 *hidden_states.shape[:-1], -1